See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/344544298

Determination of selenium in dietary supplements by ETAAS and HG-AAS: A

comparative study

Article in Atomic Spectroscopy · August 2002

CITATIONS READS

18 398

6 authors, including:

Liliana Valiente Emiliano Romero

Instituto Nacional de Tecnologia Industrial Cheplapharm GmbH
18 PUBLICATIONS 347 CITATIONS 7 PUBLICATIONS 75 CITATIONS

SEE PROFILE SEE PROFILE

Patricia Smichowski Atomic Spectroscopy

National Scientific and Technical Research Council Atomic Spectroscopy
143 PUBLICATIONS 4,046 CITATIONS 765 PUBLICATIONS 5,032 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Atomic Spectroscopy on 08 October 2020.

The user has requested enhancement of the downloaded file.

 Determination of Selenium in Dietary Supplements
by ETAAS and HG-AAS: A Comparative Study
Liliana Valientea, Margarita Piccinnaa, Emiliano Romero Alea, Ariel Grilloa, and *Patricia Smichowskib
a Instituto Nacional de Tecnología Industrial, CEQUIPE

Laboratorio de Análisis de Trazas

Casilla de Correo 157, B1650KNA-San Martín, Pcia. Buenos Aires, Argentina
b Comisión Nacional de Energía Atómica, Centro Atómico Constituyentes, Unidad de Actividad Química

Av. Gral. Paz 1499, B1650KNA-San Martín, Pcia. Buenos Aires, Argentina

INTRODUCTION within a pre-established

ABSTRACT concentration range.
Selenium has long been known A study was carried out to
as a necessary component in the establish a reliable procedure Dietary supplements of Se have
human diet and its occurrence and for determining Se in dietary become very popular due to its
function in biological systems has supplements. Three different documented anti-carcinogenic
been well-documented (1). brands of Se food supplements activity and beneficial effects on
were evaluated. These cerebrovascular diseases (7).
Selenium is an essential nutrient supplements were digested Selenium also plays an antioxidant
for humans. Food is the main with HNO3 and H2O2 using a role because it protects the cell
source for incorporation into the two-cycle microwave procedure. membrane from the pernicious
human body, and drinking water is Electrothermal atomic absorption action of peroxides. For these
responsible for a significant fraction spectrometry (ETAAS) and
reasons, it is of particular
of the total intake (2). Amounts of hydride generation-atomic
absorption spectrometry importance to evaluate the Se
55–70 µg of Se have been content in dietary products.
recommended as the daily intake (3). (HG-AAS) techniques were
selected for total Se determination. B’Hymer and Caruso (8) analyzed
Daily consumption of less than 0.1 yeast-based Se supplements using
The adverse effects of potentially
mg kg-1 per body weight will result interfering metals that are high performance liquid
in Se deficiency while levels above normally present in these chromatography (HPLC) and
1 mg kg-1 are considered toxic products were taken into ICP-MS. The authors reported that
(4,5). Deficiency of Se has been account. In particular, the effect the total Se levels were between
associated with an endemic form of Cu, Mn, and Zn on the Se 91 and 111% of label value, while
of cardiomyopathy. A variety of signal was investigated. Copper
the content of selenomethionine
other diseases, including cancer, reduces the Se signal
significantly when HG-AAS was was often found to be inaccurate.
have also been linked to a low Se
status (6). The maximum allowed used for quantification. On the Various clients of our analytical
other hand, no interference service group (CEQUIPE, INTI)
concentration of Se in drinking
effects were detected when
waters recommended by the WHO asked for total Se determination
ETAAS was used and for this
(World Health Organization) is reason it was selected as the in different types of dietary
10 µg L-1. Selenium concentrations best alternative to analyze the supplements. According to
higher than 130 µg L-1 are toxic. samples. Argentinian regulations, the Se
The overall approach was content in imported products must
Atomic absorption spectrometry tested in tablets of vitamins- be controlled. For this reason, a
(AAS) [in conjunction with hydride minerals-amino acids, nutritional reliable and robust methodology
generation (HG-AAS)], electrothermal supplements and Se-enriched for total Se determination is
atomic absorption spectrometry yeast. The selenium content required.
(ETAAS), and inductively coupled ranged from 15.9±0.7 to
plasma mass spectrometry (ICP-MS) 81.1±3.3 µg of Se per tablet. The In this study, the two analytical
have gradually assumed prime data obtained demonstrate that techniques ETAAS and HG-AAS are
importance in the ability to verify the Se content reported on the compared in order to establish a
whether foodstuffs and labels of dietary supplements are reliable methodology for the
often inaccurate. The detection determination of total Se in dietary
pharmaceutical products comply
limits were found to be 0.18 µg
with health requirements and the supplements.
g-1 Se and 0.06 µg g-1 Se for
national or international ETAAS and HG-AAS, respectively.
regulations. Specifically for Se, it is Precision was always better than
important to assess whether it is 6%, while the recovery data
ranged between 93 and 105%.
*Corresponding author.
e-mail: smichows@cnea.gov.ar

Atomic Spectroscopy 129

Vol. 23(4), July/August 2002
EXPERIMENTAL operating temperatures of 200ºC additional purification by sub-boiling
and operating pressure of distillation in a quartz still.
Instrumentation 1.38 MPa, were used for
dissolution of the samples. The mixed Pd and Mg(NO3)2
A PerkinElmer Model 5100 ZL matrix modifier solution was
atomic absorption spectrometer Reagents and Standard Solutions prepared by adding 5 mL of 0.3%
(PerkinElmer, Shelton, CT, USA), Mg(NO3)2 solution and 2.5 mL of
equipped with a PerkinElmer All chemicals were of analytical 10 g L-1 Pd solution (Merck) into a
Model THGA graphite furnace, reagent grade unless otherwise 25-mL volumetric flask, then
PerkinElmer Model AS-71 stated. Deionized water bringing to volume with deionized
autosampler, and longitudinal (Barnstead, Dubuque, IA, USA) water. The final concentration of
Zeeman-effect background was used throughout. All solutions the matrix modifier solution was
corrector, was used for the atomic were stored in high-density 0.06% Mg(NO3)2 and 0.1% Pd.
absorption measurements. polypropylene bottles. Plastic High-purity Ar was used to purge
Electrodeless discharge lamps (EDL, bottles, autosampler cups, and air from the graphite tubes.
PerkinElmer) were the sources of glassware were cleaned by
radiation used for Se determination. soaking in 20% (v/v) HNO3 for 24 Samples and Sample Handling
Pyrolytically coated graphite tubes h, then rinsed three times with
deionized water. Commercially Three brands (A, B, and C) of
with pyrolytic graphite L´vov dietary supplements were analyzed
platforms were employed. available 1000 mg L-1 Se standard
solutions (Merck, Darmstadt, for total Se content. Content
High-purity Ar (flow rate: 250 mL uniformity testing was performed
min-1) was used to purge air from Germany) were prepared daily
by serial dilution of the stock by randomly selecting five tablets
the graphite tubes, except during from each bottle. The tablets of
the atomization step where solutions. Analytical reagent nitric
acid (Merck) was used after each batch were ground and mixed
stopped flow conditions were used. thoroughly using a mortar and
The analytical measurements were
based on peak area. Autosampler
volumes of 20 µL of sample, TABLE I
followed by 5 µL of chemical Operating Conditions for Se Determination
modifier were employed for all in Dietary Supplements by ETAAS
studies. Each analysis was repeated Analytical wavelength Se: 196.0 nm
at least three times to obtain the Slit width 2.0 nm
average value and its relative Electrodeless discharge lamp 210 mA
standard deviation (%RSD). The
ETAAS operating conditions and Injection volume (sample) 20 µL
matrix modifier used are Injection volume (modifier) 5 µL
summarized in Table I. Matrix modifier Pd 0.10 % + Mg(NO3)2 0.06 % (m/v)
Measurement mode Peak area
An automated HG-AAS system
Purge gas (flow rate) 250 mL min-1 (Ar)
based on a PerkinElmer 4000
atomic absorption spectrometer,
equipped with a PerkinElmer TABLE II
MHS-20 mercury/hydride system, Instrumental Parameters and Hydride Generation Conditions
was used for the hydride for Se Determination by HG-AAS
generation studies. Electrodeless
discharge lamps (EDL, PerkinElmer) Instrument PerkinElmer 4000 AAS
were the sources of radiation used Analytical wavelength 196.0 nm
for Se determinations. The Slit width 2.0 nm
instrumental parameters of the Electrodeless discharge lamp 280 mA
HG-AAS system and the hydride Measurement mode Peak height
generation conditions for Se
determination are listed in Table II.
Hydride generation conditions
A CEM MSD-2000 microwave Instrument PerkinElmer MHS-20
digestion system (CEM, Matthews, Sample acidity 5 mol L-1 HCl
NC, USA) with a maximum
NaBH4 concentration 3% (in 1% NaOH)
microwave power of 630 W,
Cell temperature 900°C

130
 Vol. 23(4), July/August 2002

pestle. Three sub-samples of each TABLE III

mixture were accurately weighed Microwave Digestion Program
and used for the digestion step.
Sample weight 0.50 g
Sample Treatment Reagents 10 mL HNO3 (70%)
Selenium quantification by 2 mL H2O2 (30%) added in the second cycle
HG-AAS and ETAAS requires Final volume 100 mL
preliminary digestion of the Microwave program:
samples into liquid solutions. Two First cycle Steps
sample treatment procedures Power (%) 100 100 100 100
were followed depending on the
technique used for subsequent Se Pressure (psi) 40 85 135 170
quantification. Time (min) 10 10 6 6
All experiments were performed
in triplicate. In all cases, a set of Second cycle
digestion blanks was prepared. In Power 100%
order to decrease digestion blanks, Pressure 80 psi
purified nitric acid was used. Time 10 min
Procedure 1
(Se Determination by ETAAS)
A 0.5-g sample was transferred
into Teflon® vessels and 10 mL of Procedure 2 no standard addition calibration
purified concentrated HNO3 (Se Determination by HG-AAS) was necessary. The calibration was
(Merck, 70%) and 2.0 mL of H2O2 After microwave treatment (as checked every 20 measurements
(Merck, 30%) were added. The described above), the vessels were with analyte solutions varying from
average microwave (MW) allowed to cool, then opened, 10 to 100 ng mL-1.
pressure applied during the placed on a hot plate, and
digestion steps varied from 40 evaporated to dryness. The residue Analytical Determinations
to 170 psi and the complete was dissolved by dropwise addition The hydride generation in
mineralization cycle was less than of 5 mL (1+1) HCl solution. The conjunction with AAS was
45 min. The operating conditions resulting solution was placed in a performed which enables the
for MW digestion of the dietary boiling water bath (loosely covered detection level for Se to be reduced
supplements are listed in Table III. with a watch glass) and heated at by at least two orders of magnitude.
Optimization of power and time in 950C for about 15 min to convert
all Se to Se(IV). The digest was Selenium concentrations were
the MW digestion is important
allowed to cool and was then also determined by ETAAS by
in order to achieve complete
transferred into a 100-mL injecting 20-µL aliquots of digested
dissolution of the samples. The
volumetric flask. Eight mL of sample and 5 µL of chemical matrix
digestion solutions were allowed
concentrated HCl was added and modifier at each firing into the
to cool, transferred into 100-mL
the solution diluted to volume pyrolytically coated tubes and
graduated flasks, and diluted to the
with deionized water. applying the operating conditions
mark with deionized water.
and the temperature program given
For assessing the accuracy of For assessing the accuracy of the in Tables I and IV, respectively.
the method, aliquots of the method, aliquots of CRM TORT-1, The signals were registered as
CRM TORT-1, Lobster hepatopan- Lobster hepatopancreas (NRCC, peak area.
creas (NRCC, Ottawa, Québec, Ottawa, Québec, Canada) were
treated in the same manner as Unfortunately, and to the best
Canada) were treated in the same
the samples. of our knowledge, no CRMs for
manner as the samples.
Se determination in dietary
Calibration supplements are currently available.
For this reason, TORT-1 was
The calibration curves were chosen as a possible alternative to
obtained with calibration standards check the accuracy of the method.
prepared in the same acid medium In addition, a recovery test
as the samples. In spite of the was performed.
complexity of the matrix analyzed,

131
 TABLE IV performed at high temperatures
Graphite Furnace Temperature Program (90–100ºC) since the redox
Parameter Drying Pyrolysis Atomi-- Con- equilibrium is temperature-
Step 1 Step 2 zation ditioning dependent (11). To convert Se(VI)
to Se(IV) for maximum efficiency in
Temperature (°C) 110 130 1100 2150 2400 the generation of SeH2, the digested
Ramp time (s) 1 15 10 0 1 samples were heated at 95ºC for 15
min in (1+1) HCl solution.
Hold time (s) 30 30 20 5 2
The effect of tetrahydroborate
Ar flow rate (mL min-1) 250 250 250 0 (read) 250 (III) concentration on the Se
response was also examined.
Optimal Se signals were achieved
with 3% (w/v) NaBH4. For
concentrations higher than 3.5%,
the sensitivity decreased slightly.
The optimized chemical and
physical hydride generation
parameters are listed in Table II.
The generation of SeH2 is very
fast and is estimated to take place
in less than 10 ms (12). The
generation conditions were
optimized taking special
consideration of the kinetics of
the reaction since Se is normally
reduced faster than the interfering
transition metals. A rapid separation
of the gaseous products was also
taken into account.
Three typical transition metals
present in dietary supplements
were tested in order to evaluate if
the Se signal is affected in their
presence when optimized
Fig. 1. Interference study. conditions are used to generate the
Se hydride. The concentrations of
Cu, Mn, and Zn listed on the labels
RESULTS AND DISCUSSION increases with higher power. The of brands A and C were 1 mg, 5 mg,
addition of peroxide to the nitric and 20 mg, respectively. The labels
Selection of Digestion Reagents acid was necessary to increase the of brand B listed 2 mg Cu, 2.5 mg
The selection of the type and oxidation efficiency. Mn, and 15 mg Zn.
amount of acid concentration and
MW operating parameters is HG-AAS vs. ETAAS to Determine All solutions analyzed contained
important for the successful Se in Dietary Supplements 60 µg L-1 of Se and the results are
dissolution of the samples. Nitric the average of three measurements.
Owing to the high oxidation
acid was used due to its excellent The amounts of Cu, Mn, and Zn
potential of acid mixtures, most
dissolving capabilities. Preliminary added represent 30, 70, and 100% of
parts of the selenium species are
studies demonstrated that a the content of each metal in the
probably converted into selenate.
maximum power last-stage was dietary supplements according to
A pre-reduction step for converting
beneficial for better digestion label claims. The maximum
the selenium species into Se(IV) is
(9,10). This is attributed to the concentration of the metals in the
necessary in all HG procedures
fact that the rate of the digestion final solution were 6 µg mL-1 Cu;
because only Se(IV) is capable of
reaction of acid decomposition and 30 µg mL-1 Mn, and 120 µg mL-1 Zn.
generating the Se hydride. This
oxidation power of the acids Figure 1 shows that the interference
reduction step should be

132
 Vol. 23(4), July/August 2002

effect caused by Cu is serious to the Quality Parameters reported by the manufacturer.

extent that the Se The detection and Total Se levels found were between
signal was significantly reduced quantification limits were 10.0±0.5 and 81.1±3.3 µg per
when 1.8 µg mL-1 of Cu was added. calculated following the IUPAC tablet based on a five-tablet
Reductions between 25 and 40% regulations based on the 3  and composite assay. The label listed
were also observed when Mn and 10  criterion, respectively, for 10 20 µg Se per tablet for brands
Zn were tested at concentrations of replicate measurements of the A and C and 25 µg Se per tablet
9.0 and 36 µg mL-1, respectively blank signal. The relative standard for brand B. It is important to note
deviation (%RSD) was calculated that significant differences were
In an effort to overcome the found between bottles of the same
effect observed, the reductant for 10 successive measurements
of a solution containing a final trademark. Brand A exhibited a
concentration was decreased and variation among bottles ranging
the concentration of HCl increased. concentration of 50 ng mL-1 Se.
The results are summarized in from 15.9±0.6 to 27.2±1.1 µg of Se
Studies on the mechanism of per tablet. The variation in brand B
transition metal interferences in Table V together with the results of
the accuracy test using ETAAS and was between 25.2±1.2 and
HG-AAS have demonstrated that 81.1±3.3 µg of Se per tablet,
smaller amounts of reductant can HG-AAS.
and in brand C from 10.0±0.5 to
substantially improve the tolerance Unfortunately, no CRMs for 29.5±1.3. Brand B exhibited a
to transition metal interferences Se in the studied matrix are difference of over 300%.
(13). Some authors prefer to use commercially available and for this
higher HCl concentrations in order reason, a recovery test was CONCLUSION
to increase the efficiency of SeH2 performed on several dietary
production (14,15). However, we supplements. Samples 1-A, 6-A, 1-B, The data obtained in this study
found that under the experimental 5-B, 1-C, and 2-C were spiked with show that the major shortcoming
conditions of this study, no 20 µg of Se and the over-all of the HG-AAS approach is that the
significant improvement in Se analytical procedure followed. The presence of transition metals such
recovery was observed. recovery data ranged from 93 to as Cu, Mn and Zn affects the
105%. Although this cannot replace formation and release of SeH2,
The situation was different when with Cu causing the most serious
ETAAS was used to analyze the an accuracy test, the information
gained verifies the good interference.
samples and no matrix effects were
observed. Figure 1 shows that Cu, performance of the method. On the other hand, ETAAS
Mn, and Zn do not affect the provides an excellent means for
Se signal. Based on these results, Determination of Se in Selected the determination of total Se in
ETAAS was chosen as the best Dietary Supplements dietary supplements.
alternative to analyze dietary The Se content was determined
by ETAAS in three different This study also shows that the
supplement samples.
commercial dietary supplements. Se content reported on labels of
Table VI lists the Se concentrations dietary supplements is often
found vs. the content per tablet inaccurate.

Received July 16, 2002.

TABLE V
Performance Levels for the Se Determination
in Dietary Supplements Using ETAAS and HG-AAS
Parameter ETAAS HG-AAS
Detection limita 0.18 µg g-1 0.06 µg g-1
Quantification limita 0.6 µg g-1 0.2 µg g-1
Precisiona 3.7% 5.6%
Accuracy testb
- Certified value 6.88±0.47 µg g-1 6.88±0.47 µg g-1
- Found 6.6 ± 0.3 µg g-1 6.1±0.4 µg g-1
a
n=10.
b
CRM TORT-1, Lobster hepatopancreas (NRCC, Ottawa, Québec, Canada).

133
 TABLE VI REFERENCES
Selenium Content in Commercial Dietary Supplements 1. A. Stacchini, E. Conti, M. Baldini, E.
Results expressed in µg of Se per tablet. Beccaloni, and S. Caroli, J. Trace
Sample a
Type of supplement Se reported in Se found Elem. Electrolytes Health Dis. 3,
prospectus 193 (1989).
1-A Vitamins-minerals-aminoacids 20 20.0±0.9 2. E. J. Underwood, Trace elements in
human and animal nutrition,
2-A Vitamins-minerals-aminoacids 20 21.1±0.8 Academic Press, New York (1971).
3-A Vitamins-minerals-aminoacids 20 22.1±0.9 3. J. L. Gómez-Ariza, J. A. Pozas, I.
Giráldez , and E. Morales, Analyst
4-A Vitamins-minerals-aminoacids 20 27.2±1.1 124, 75 (1999).
5-A Vitamins-minerals-aminoacids 20 17.9±0.7 4. O. Wada, N. Kurihara, and N.
Yamazaki, Jap. J. Nutr. Assess. 11,
6-A Vitamins-minerals-aminoacids 20 18.0±0.8 48 (1995).
7-A Vitamins-minerals-aminoacids 20 15.9±0.6 5. O. A. Levander, Ann. Rev. Nutr. 7,
227 (1987).
1-B Nutritional supplement 25 81.1±3.3 6. L. Fishbein, Metals and their
2-B Nutritional supplement 25 31.1±1.1 compounds in the environment,
VCH, Germany (1991).
3-B Nutritional supplement 25 35.0±1.1 7. A. M. Fang and K. W. Kizer, West J.
4-B Nutritional supplement 25 27.8±1.0 Med. 153, 160 (1990).
8. C. B´Hymer and J. A. Caruso, J. Anal.
5-B Nutritional supplement 25 25.2±1.2 At. Spectrom. 15, 1531 (2000).
1-C Se-enriched yeast 20 29.5±1.3 9. P. Smichowski, L. Valiente, and M.
Piccinna, At. Spectrosc. 22, 350
2-C Se-enriched yeast 20 27.2±0.9 (2001).
3-C Se-enriched yeast 20 10.0±0.5 10. J. Marrero, S. Farías, and P.
a
Smichowski, Quim. Anal. 20, 13
A, B, C are different commercial products and the figures represent different samples (2001).
of the same trademark.
11. M. Verlinden, H. A. Deelstra, and E.
Adriaensses, Talanta 28, 6 (1981).
12. J. Agterdenbos, and D. Bax, Anal.
Chim. Acta 188, 127 (1986).
13. B. Welz, and M. Schubert-Jacobs, J.
Anal. At. Spectrom. 1, 23 (1986).
14. B. Welz, Chem. Br. 22, 130 (1986).
15. M. Yamamoto, M. Yasuda, and Y.
Yamamoto, Anal. Chem. 57, 1382
(1985).

134

View publication stats